用R的rugarch package：
The rugarch package is the premier open source software for univariate GARCH modelling. It is written in R using S4 methods and classes with a significant part of the code in C and C++ for speed. It contains a number of GARCH models beyond the vanilla version including IGARCH, EGARCH, GJR, APARCH, FGARCH, Component-GARCH, multiplicative Component-GARCH for high frequency returns and the realized-GARCH model, as well as a very large number of conditional distributions including (Skew)-Normal, (Skew)-GED, (Skew)-Student (Fernandez/Steel), (Skew)-Student (GH), Normal Inverse Gaussian (NIG), Generalized Hyperbolic (GH) and Johnson’s SU (JSU). The conditional mean equation includes ARFIMA and ARCH-in-mean, and is estimated in a joint step with the GARCH model. Both the conditional mean and variance parts allow for external regressors to be used. A comprehensive set of methods to work with these models are implemented, and include estimation, filtering, forecasting, simulation, inference tests and plots, with additional functionality in the form of the GARCH bootstrap, parameter uncertainty via the GARCH distribution function, misspecification tests (Hansen’s GMM and Hong&Li Portmanteau type test), predictive accuracy tests (Pesaran&Timmermann, Anatolyev&Gerko), and Value at Risk tests (VaR Exceedances and Expected Shortfall tests).

下载：
https://cran.r-project.org/web/packages/rugarch/index.html

首先感谢热心提供，然后我要说的是，这里面也只包含了realized-GARCH model，而并没有realized-EGARCH model，虽然只是取了对数，但是并不能直接使用，所以我还要研究一下才能给分。

问下那你现在改好了吗…………我也在做这个东西，所以也想找下相关的程序包……

同求改好的程序

同求改好的程序